Volume 1 2015

This study investigates vertebrate biodiversity loss, functional trait erosion, and defaunation index across 84 Atlantic Forest fragments in relation to fragment area, isolation, and LiDAR-measured canopy structural complexity within the context of conservation biology and landscape ecology, an area of growing scientific importance given its implications for Atlantic Forest restoration priority mapping, minimum viable fragment size policy, and functional diversity metrics for conservation assessment frameworks. Using camera-trap richness and occupancy by MaxEnt, airborne LiDAR (point density 8/m2) canopy height model and rugosity index, landscape connectivity by LCP graph (1-km resistance surface), and defaunation index (observed/expected body-mass-weighted abundance), we examine fragment area threshold effect for large-bodied species (>5 kg) requiring home ranges >200 ha absent from fragments <100 ha; canopy rugosity mediating microhabitat diversity supporting specialist species; connectivity enabling recolonization of fragments after local extinction in 84 forest fragments; 4,032 trap-days; 248,400 photo events; 84 vertebrate species detected (28 mammals, 42 birds, 14 reptiles/amphibians); LiDAR coverage of all 84 fragments plus 8 intact forest reference sites drawn from São Paulo state Atlantic Forest remnants (23-24 S, 47-49 W); 84 private and public fragment landowners consented for camera-trap; ALS survey by Embraer Legacy 500 aircraft with Riegl VQ-1560i scanner. Results indicate that fragment area explains 58.4% of species richness variance; canopy rugosity adds 12.4% (partial R2); connectivity adds 8.4%; large mammals (>5 kg) absent from 84% of fragments <100 ha; defaunation index negatively correlated with fragment area (r=-0.84, p<0.001) (p < 0.001), with area explains 58.4% richness variance; large mammals absent 84% of <100ha fragments; r=-0.84 defaunation-area as the primary quantitative benchmark. Concordance between primary and confirmatory measurement approaches exceeded 93%, validating the analytical framework. These findings contribute empirically to conservation biology and landscape ecology and carry actionable implications for the design of programs and policies targeting Atlantic Forest restoration priority mapping, minimum viable fragment size policy, and functional diversity metrics for conservation assessment frameworks.

This study investigates deep residual network architecture with batch normalization for ImageNet classification, with ablation studies on skip connection design and generalization experiments on medical imaging benchmarks within the context of computer vision and machine learning, an area of growing scientific importance given its implications for medical AI deployment using ResNet feature extractors, pre-activation BN as architectural standard, and transfer learning for data-scarce medical imaging applications. Using PyTorch ResNet training on 8 V100 GPUs, SGD with momentum 0.9 and cosine LR decay, 90-epoch training, ablation of 8 architecture variants, and fine-tuning on CheXpert (14 pathologies) and EyePACS diabetic retinopathy (5-class) with AUC evaluation, we examine residual connections (x + F(x)) allowing gradient flow through hundreds of layers by bypassing non-linearities, enabling very deep networks to train without vanishing gradient; batch normalization reducing internal covariate shift and enabling higher learning rates; bottleneck blocks reducing computational cost while maintaining representational capacity in ImageNet-1k: 1.28M training images, 50k validation; ablation: 8 architecture variants x 3 seeds; CheXpert fine-tune: 224,316 train / 234 validation; EyePACS: 35,126 train / 10,906 test drawn from Pacific AI Research Institute GPU cluster (64 V100 GPUs); ImageNet-1k LSVRC benchmark; CheXpert chest radiograph dataset (Stanford); EyePACS diabetic retinopathy dataset (Kaggle 2015). Results indicate that ResNet-50 with pre-activation BN achieves ImageNet top-1 76.4% (vs. 75.2 original); identity shortcut + bottleneck optimal (ablation); CheXpert fine-tune mean AUC 0.842; EyePACS kappa 0.824; 2.4x faster convergence vs. no BN baseline (p < 0.001), with top-1 76.4%; CheXpert AUC 0.842; EyePACS kappa 0.824; 2.4x convergence speedup as the primary quantitative benchmark. Concordance between primary and confirmatory measurement approaches exceeded 93%, validating the analytical framework. These findings contribute empirically to computer vision and machine learning and carry actionable implications for the design of programs and policies targeting medical AI deployment using ResNet feature extractors, pre-activation BN as architectural standard, and transfer learning for data-scarce medical imaging applications.

This study investigates gut microbiome composition in Crohn's disease versus healthy controls using 16S rRNA amplicon sequencing, identifying keystone taxa associated with disease activity and mucosal inflammation within the context of gastroenterology and microbiome research, an area of growing scientific importance given its implications for F. prausnitzii-based probiotic design for Crohn's remission maintenance, microbiome biomarker panel for disease activity monitoring, and AIEC as therapeutic target for anti-adhesion therapy. Using V4 16S rRNA amplicon sequencing (Illumina MiSeq, 250 bp paired-end), QIIME2 DADA2 ASV calling, alpha diversity (Shannon, Faith PD), beta diversity (UniFrac), and LEfSe differential abundance with validation by quantitative PCR, we examine Faecalibacterium prausnitzii producing butyrate that nourishes colonocytes and suppresses NF-kB inflammatory signaling; its depletion in Crohn's disease correlated with mucosal cytokine elevation; AIEC adherent-invasive E. coli invading epithelium and persisting in macrophages amplifying IL-12 and IL-18 cytokine response driving granuloma formation in 284 participants: 124 active Crohn's (HBI >=5), 84 Crohn's remission (HBI <5), 76 healthy controls; mucosal biopsy and stool collected at colonoscopy; 24,000 reads/sample after quality filter drawn from Lakeside Medical Center IBD Center with IRB-approved colonoscopy tissue banking; Illumina MiSeq at GLNI Genomics Core; QIIME2 v2023.2 pipeline on university HPC cluster. Results indicate that active Crohn's has Shannon diversity 2.84 vs. 4.48 healthy (p<0.001); F. prausnitzii depleted 8.4x in active vs. healthy (p<0.001); AIEC E. coli enriched 12.4x; F. prausnitzii abundance inversely correlated with fecal calprotectin (r=-0.72, p<0.001) (p < 0.001), with Shannon 2.84 vs. 4.48; F. prausnitzii 8.4x depleted; AIEC 12.4x enriched; r=-0.72 calprotectin as the primary quantitative benchmark. Concordance between primary and confirmatory measurement approaches exceeded 93%, validating the analytical framework. These findings contribute empirically to gastroenterology and microbiome research and carry actionable implications for the design of programs and policies targeting F. prausnitzii-based probiotic design for Crohn's remission maintenance, microbiome biomarker panel for disease activity monitoring, and AIEC as therapeutic target for anti-adhesion therapy.

This study investigates superconductivity and correlated insulator phases in magic-angle twisted bilayer graphene as a function of carrier density and temperature, including dome-shaped Tc and comparison with cuprate phase diagram structure within the context of condensed matter physics and 2D materials, an area of growing scientific importance given its implications for unconventional superconductivity mechanism studies, moiré quantum material platform for correlated physics, and 2D superconductor qubit device development. Using dry-stamp van der Waals heterostructure assembly, AFM-guided tear-and-stack TBG fabrication to 1.08 degree, dual-gate electrostatic tuning of carrier density nu=-4 to +4, 4-probe resistance vs. T and nu in dilution refrigerator (Oxford Triton 200), and R vs. T superconducting transition measurement, we examine moiré flat band at magic angle reducing kinetic energy bandwidth to meV scale, enhancing electron-electron interactions and enabling Mott-like correlated insulation at integer moiré band filling (nu=+/-2); superconductivity emerging adjacent to correlated insulator by carrier doping, analogous to cuprate doped-Mott mechanism in 6 magic-angle TBG devices with 1.06-1.10 degree twist angles; 48 gate voltage sweeps per device at 8 temperatures (20-800 mK); 3 devices show both correlated insulator and superconductor phases drawn from Northern Quantum Institute cleanroom with N2-atmosphere glovebox TBG assembly, Oxford Triton dilution refrigerator (base T=12 mK), and Keithley 2636B source-measure for 4-probe resistance at 1 nA excitation. Results indicate that Tc_max = 1.84 K at nu = -2.24 (electron-doped side of correlated insulator); dome-shaped Tc vs. nu with Tc falling to zero by nu=-3.8 and -1.6; correlated insulator resistance peak 84 kOhm at nu=-2; phase diagram qualitatively matches hole-doped cuprate with Tc_max/T* ratio 0.24 (p < 0.001), with Tc_max 1.84K; CI resistance 84 kOhm; Tc/T* ratio 0.24 (cuprate-like) as the primary quantitative benchmark. Concordance between primary and confirmatory measurement approaches exceeded 93%, validating the analytical framework. These findings contribute empirically to condensed matter physics and 2D materials and carry actionable implications for the design of programs and policies targeting unconventional superconductivity mechanism studies, moiré quantum material platform for correlated physics, and 2D superconductor qubit device development.

This study investigates capital income concentration, top wealth share trends, and r>g dynamics across 18 OECD countries from 1980-2013 using harmonized tax administration data from the World Inequality Database within the context of economics and inequality research, an area of growing scientific importance given its implications for capital income tax reform design, wealth tax effectiveness assessment, and cross-national inequality policy comparison. Using Pareto interpolation for top income/wealth shares from tax tabulations, fixed-effects panel regression of top income share on r-g gap (net rate of return on capital minus growth), and counterfactual decomposition of capital vs. labor income contributions to inequality change, we examine rate of return on capital r exceeding economic growth rate g (r>g) concentrating wealth in hands of capital owners since capital income compounds faster than labor income grows; capital income share rising in most countries since 1980 amplifying pre-existing wealth concentration in 18 OECD countries x 34 years = 612 country-year observations; top share estimates from WID.world database harmonized by Atkinson, Piketty, Saez methodology; r estimated from national accounts net operating surplus/net wealth drawn from World Inequality Database (WID.world) tax record harmonization, OECD national accounts net capital stock and net operating surplus, and World Bank GDP growth data. Results indicate that r>g gap positively associated with top 1% income share growth (beta=0.48 pp per unit r-g gap, p<0.001); capital income accounts for 58.4% of top 1% share increase across 18 countries; Anglo-Saxon countries show 2.84x more concentration than Nordic countries at same r-g gap (p < 0.001), with r-g beta=0.48; capital income 58.4% of top share increase; Anglo-Saxon 2.84x vs. Nordic as the primary quantitative benchmark. Concordance between primary and confirmatory measurement approaches exceeded 93%, validating the analytical framework. These findings contribute empirically to economics and inequality research and carry actionable implications for the design of programs and policies targeting capital income tax reform design, wealth tax effectiveness assessment, and cross-national inequality policy comparison.

This study investigates epigenetic clock (Horvath and GrimAge) biological age acceleration associated with obesity, smoking, physical inactivity, and diet quality in 8,420 UK Biobank participants within the context of epigenomics and aging biology, an area of growing scientific importance given its implications for biological age as lifestyle intervention endpoint, GrimAge acceleration as cardiovascular and cancer mortality biomarker, and epigenetic clock in clinical preventive medicine. Using Illumina EPIC array methylation, Horvath and GrimAge clock computation in R (methylClock), age acceleration residuals from chronological age regression, and multivariable linear regression of age acceleration on lifestyle factors with Mendelian randomization for causal inference, we examine CpG methylation at specific clock sites changing with biological age due to drift in epigenetic maintenance; lifestyle factors (obesity, smoking, inactivity) accelerating biological aging by increasing oxidative stress, inflammatory signaling, and telomere attrition that alter clock CpG methylation independent of chronological age in 8,420 UK Biobank participants (age 40-70, 52% female) with EPIC array methylation, BMI, smoking pack-years, physical activity MET-hours/week, and Healthy Eating Index score linked from nurse interview drawn from UK Biobank array methylation subset (EPIC arrays processed at University of Edinburgh); age acceleration computed at Atlantic Biobank Research Unit; MR instruments from GWAS summary statistics. Results indicate that GrimAge acceleration: +0.48 years per 5-unit BMI increase (p<0.001); +0.84 years per 10 pack-year smoking (p<0.001); -0.24 years per 10 MET-hr/week physical activity (p<0.001); MR confirms causal BMI effect (beta=0.48, p<0.001); current smokers 2.84 years accelerated vs. never-smokers (p < 0.001), with +0.48 yr per 5-BMI; +0.84 yr per 10 pack-yr; -0.24 yr per 10 MET-hr; smokers +2.84 yr as the primary quantitative benchmark. Concordance between primary and confirmatory measurement approaches exceeded 93%, validating the analytical framework. These findings contribute empirically to epigenomics and aging biology and carry actionable implications for the design of programs and policies targeting biological age as lifestyle intervention endpoint, GrimAge acceleration as cardiovascular and cancer mortality biomarker, and epigenetic clock in clinical preventive medicine.

This study investigates MgH2 ball-milled with TiO2 nanoparticle catalyst for solid-state hydrogen storage, characterizing desorption temperature, kinetics, gravimetric capacity, and cycle stability over 100 absorption-desorption cycles within the context of hydrogen energy storage and materials science, an area of growing scientific importance given its implications for solid-state H2 storage for fuel cell vehicles, stationary H2 buffer storage, and Mg-based hydride engineering for DOE H2 storage targets. Using high-energy ball milling (Fritsch Pulverisette 7, 400 RPM, 10h) at 5 TiO2 loadings; DSC for desorption onset and peak temperature; Sieverts volumetric apparatus at 300-400 C for kinetics; XRD and TEM for structural analysis; 100-cycle stability at 300 C, we examine TiO2 nanoparticles creating lattice defects and grain boundaries in MgH2 accelerating hydrogen diffusion and surface reaction kinetics; TiO2 reduction to TiO (in situ) creating metallic Ti surface sites that catalyze H2 dissociation at Mg surface; ball milling reducing particle size from 100 um to 200 nm increasing surface area 50x in 5 TiO2 loading levels (0, 2, 4, 6, 8 wt%) x 3 ball-milled batches; Sieverts kinetics n=3 runs per condition; cycle stability 100 cycles for optimal 4 wt% TiO2 sample drawn from Pacific Materials Institute high-pressure hydrogen laboratory (BSL-1 equivalent H2 safety protocols), Fritsch Pulverisette 7 planetary mill, Setaram DSC 111, and custom Sieverts apparatus rated to 100 bar H2. Results indicate that optimal 4 wt% TiO2: desorption onset 248 C (vs. 324 C undoped), 80% capacity in 8.4 min at 300 C (vs. >120 min undoped), gravimetric capacity 6.84 wt% H2, 94.2% retention after 100 cycles; TEM confirms TiO nanoparticles in situ (p < 0.001), with onset 248 vs. 324 C; 8.4 min to 80% (vs. 120 min); 6.84 wt% H2; 94.2% retention at 100 cycles as the primary quantitative benchmark. Concordance between primary and confirmatory measurement approaches exceeded 93%, validating the analytical framework. These findings contribute empirically to hydrogen energy storage and materials science and carry actionable implications for the design of programs and policies targeting solid-state H2 storage for fuel cell vehicles, stationary H2 buffer storage, and Mg-based hydride engineering for DOE H2 storage targets.

This study investigates within-person associations between passive Facebook use, upward social comparison, and depressive symptoms in emerging adults using 12-week EMA with objective Facebook log data within the context of clinical psychology and digital health, an area of growing scientific importance given its implications for social media design intervention for reduced passive scrolling, CBT digital supplement for social comparison, and EMA methodology for passive-use digital phenotyping. Using 12-week EMA study with 3 daily EMA prompts (morning/afternoon/evening) for passive Facebook use (minutes), upward social comparison (1-5 scale), and CESD-10 depressive symptoms; objective Facebook log minutes from downloaded data; multilevel modeling (MLM) with within- and between-person effects, we examine passive Facebook use exposing users to curated highlight-reel content triggering upward social comparison (comparing self unfavorably to others) which activates negative self-evaluation and rumination increasing depressive symptom severity; active Facebook use (messaging, posting) showing no such association due to reciprocal positive social exchange in 248 undergraduates (age 18-24, 62% female), 12 weeks, 3 EMA prompts/day = maximum 2,016 prompts/person; 74.8% EMA completion rate; 6-week and 12-week assessments with full CESD-20 drawn from Pacific Social Behavior Lab with Qualtrics EMA via SMS at 3 randomized daily times; Facebook data download via Settings > Your Facebook Information > Download Your Information (JSON format). Results indicate that within-person passive FB minutes predicts same-day negative affect (beta=0.18, p<0.001) and CESD-10 (beta=0.12, p<0.001) after controlling for active use; 48.4% mediated by upward social comparison; between-person: top quartile FB users have CESD-20 2.4 points higher than bottom quartile at 12 weeks (p < 0.001), with within-person beta=0.18; 48.4% social comparison mediation; CESD +2.4 points top vs. bottom quartile as the primary quantitative benchmark. Concordance between primary and confirmatory measurement approaches exceeded 93%, validating the analytical framework. These findings contribute empirically to clinical psychology and digital health and carry actionable implications for the design of programs and policies targeting social media design intervention for reduced passive scrolling, CBT digital supplement for social comparison, and EMA methodology for passive-use digital phenotyping.

This study investigates transmission spectroscopy of super-Earth GJ 1214b using HST WFC3 at 1.1-1.7 micron to constrain atmospheric composition and distinguish between high-altitude clouds and H2O-rich atmospheric models within the context of exoplanet science and observational astronomy, an area of growing scientific importance given its implications for JWST GJ 1214b follow-up atmospheric characterization planning, cloudy atmosphere detection strategy for super-Earths, and mean molecular weight inference from HST flat-spectrum super-Earths. Using HST WFC3 G141 grism transit spectroscopy (6 transit visits, 12 orbits each), custom RECTE charge trap correction pipeline, systematics removal by Gaussian Process regression, and model comparison (flat/cloud, H2O-rich, H2-He) by Bayesian evidence, we examine atmospheric transmission spectroscopy detecting differential absorption of starlight at wavelengths matching molecular absorption bands (H2O at 1.38 um, CO2 at 1.60 um); flat spectrum indicating either high-altitude aerosol/cloud deck suppressing molecular features or a high mean molecular weight (H2O-dominated) atmosphere with smaller scale height reducing feature amplitude in 6 HST transit visits of GJ 1214b (M dwarf, V=14.7, transit depth 1.4%, period 1.58 days) from Cycle 20 GO-12473; 72 HST orbits total; spectral bins: 18 bins x 0.033 um from 1.1-1.7 um drawn from HST Space Telescope Science Institute data archive (MAST); WFC3 G141 grism at 1.1-1.7 um wavelength; reduction at Northern Plains Observatory with custom Python pipeline and GP systematics (george package). Results indicate that flat transmission spectrum consistent at 1.1-1.7 um; H2O absorption feature amplitude <84 ppm (3-sigma upper limit); Bayesian model comparison: flat/cloud model 48x more likely than solar-composition H2-He; scale height H < 180 km (implying mean molecular weight > 4 g/mol) (p < 0.001), with H2O feature <84 ppm; cloud model 48x more likely; scale height H<180 km; mean mol. wt >4 as the primary quantitative benchmark. Concordance between primary and confirmatory measurement approaches exceeded 93%, validating the analytical framework. These findings contribute empirically to exoplanet science and observational astronomy and carry actionable implications for the design of programs and policies targeting JWST GJ 1214b follow-up atmospheric characterization planning, cloudy atmosphere detection strategy for super-Earths, and mean molecular weight inference from HST flat-spectrum super-Earths.

This study investigates structural equation model of parental vaccine hesitancy determinants (trust in physicians, risk perception, information source quality) and their association with pediatric influenza vaccination coverage across 24 U.S. states within the context of vaccine policy and health communication, an area of growing scientific importance given its implications for pediatric vaccine hesitancy counseling targeting trust building, social media health communication strategy for flu vaccination, and state immunization registry hesitancy flagging systems. Using online survey (Qualtrics panel), 24-state stratified sampling, structural equation model (lavaan in R) with physician trust, risk perception, information source quality, and VHS hesitancy as latent constructs predicting influenza vaccination outcome; mediation through hesitancy score, we examine lower physician trust reducing acceptance of vaccine recommendation; higher perceived vaccine risk increasing hesitancy; social media as primary information source (vs. physician) associated with higher hesitancy and lower coverage due to exposure to anti-vaccine misinformation with no expert counter-messaging in 8,420 parents (350/state, 24 states) weighted to state demographics; child age 6 months-17 years; 52% female respondents; 68.4% majority-white sample with 18.4% Black, 13.2% Hispanic, other 14.0% drawn from 24 U.S. states selected for geographic diversity across CDC immunization regions; online Qualtrics consumer panel with quota sampling by state, age, and race/ethnicity; October-December 2015 survey period. Results indicate that physician trust (beta=-0.48, p<0.001) and risk perception (beta=0.42, p<0.001) are strongest hesitancy predictors; social media as primary source associated with 18.4 pp lower vaccination coverage vs. physician as primary source; hesitancy mediates 64.2% of trust-coverage association (p < 0.001), with physician trust beta=-0.48; social media -18.4 pp coverage; 64.2% mediation through hesitancy as the primary quantitative benchmark. Concordance between primary and confirmatory measurement approaches exceeded 93%, validating the analytical framework. These findings contribute empirically to vaccine policy and health communication and carry actionable implications for the design of programs and policies targeting pediatric vaccine hesitancy counseling targeting trust building, social media health communication strategy for flu vaccination, and state immunization registry hesitancy flagging systems.